Supported LLMs

For interactive sessions, you can directly use an LLM model from our supported LLM list:

Note: The public Interactive playground supports Gemini-series models only.

Bring Your Own LLM

You can also use a custom LLM if you have requirements on the LLM model. Our interactive model currently supports any LLMs that are served with OpenAI compatible API for chat completion, which is now widely adopted by all major inference and serving frameworks like vLLM, llama.cpp and Ollama.

If you would like to use a custom LLM, drop us a message at our Discord.